Post on 05-Aug-2020
transcript
This volume is a product of the staff of the International Bank for Reconstruction and
Development/The World Bank. The World Bank does not guarantee the accuracy of the data
included in this work. The findings, interpretations, and conclusions expressed in this paper do
not necessarily reflect the views of the Executive Directors of the World Bank or the
governments they represent.
The material in this publication is copyrighted.
FINANCIAL SECTOR ASSESSMENT PROGRAM
POLAND
HOUSING FINANCE
TECHNICAL NOTE
JANUARY 2014
THE WORLD BANK
FINANCIAL AND PRIVATE SECTOR DEVELOPMENT VICE PRESIDENCY
EUROPE AND CENTRAL ASIA REGIONAL VICE PRESIDENCY
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
1
Contents
Introduction ................................................................................................................................................... 4
Executive Summary ...................................................................................................................................... 5
Key Recommendations ................................................................................................................................. 6
Mortgage Market Context ............................................................................................................................. 7
Addressing the Risks of the Current Mortgage Portfolio .............................................................................. 9
Strengthening the Mortgage Regulatory Environment ............................................................................... 13
Modernization of the Mortgage Covered Bonds System ............................................................................ 19
Key MCB Characteristics from a Rating Agency Perspective ............................................................... 19
Approaches to Manage Asset Encumbrance ........................................................................................... 21
Covered Bonds Issuance Models ............................................................................................................ 22
Elements of the Modern Securitization Framework ................................................................................... 24
Conclusions ................................................................................................................................................. 29
Annex 1. MCB in the Legal, Regulatory and Marketplace context ............................................................ 30
Annex 2. Comparison between MCBs and RMBSs ................................................................................... 32
2
Abbreviations
ABS Asset Backed Securities
AHML Agency for Home Mortgage Lending
ARM Adjustable Rate Mortgage
BCBS Basel Committee on Banking Supervision
CBS Commission for Banking Supervision
CFS Commission for Financial Supervision
CHF Swiss Franc
CIRS Cross-currency interest rate swaps
CRD Capital Requirements Directive
CRDIV Capital Requirements Directive IV
CRR Capital Requirements Regulation
CSO Central Statistical Office
D-SIFI Domestic Systemically Important Financial Institution
DTI Debt-to-income ratio
EBA European Banking Authority
ECA Europe and Central Asia
ECB European Central Bank
ECBC European Covered Bond Council
ESCB European System of Central Banks
ESRB European Systemic Risk Board
EU European Union
EUR Euro
FNMA Federal National Mortgage Association
FRM Fixed Rate Mortgage
FSAP Financial Sector Assessment Program
FSB Financial Stability Board
FSC Financial Stability Committee
FSOC Financial Stability Oversight Council
FX Foreign exchange
GDP Gross domestic product
HPA House Price Appreciation
IFRS International Financial Reporting Standards
IT Information Technology
KNF Polish Financial Supervision Commission
LCR Liquidity Coverage Ratio
LTV Loan To Value – ratio of mortgage loan principal amount to the value of the real estate collateral
MCB Mortgage Covered Bond
MGIC Mortgage Guaranty Insurance Corporation
MoF Ministry of Finance
MSA Metropolitan Statistical Area
NBP National Bank of Poland
NPL Non-Performing Loan
OECD Organization for Economic Cooperation and Development
p.a. Per annum
PBA Polish Bankers Association
PLN Polish zloty
3
QRM Qualified Residential Mortgage
RMBS Residential Mortgage Backed Securities
UCITS Undertakings of Collective Investment in Transferable Securities (EU Directive 2009/65)
USD U.S. dollar
WSE Warsaw Stock Exchange
4
Introduction
A World Bank - IMF mission visited Warsaw February 19 – March 5, 2013, to undertake an update of the
Financial Sector Assessment Program (FSAP) conducted in 2006.1 This Note is prepared as follow-on to
the Section V (B) - Development and Market Structure (Covered Bonds) of the Aide Memoire produced
as part of that mission. The primary focus of this Note is on the development of the capital market access
mechanisms for the mortgage industry. The comprehensive approach that is suggested to the authorities
includes areas of improvements in the mortgage assets per se, legal and regulatory framework of the
mortgage loan liquidity and, finally, in the Mortgage Covered Bond (MCB) and securitization
environment.
The topic of establishing a modern MCB framework in Poland has been the focus of a number of World
Bank reports, notes and related consultations with the Polish counterparts over the last 7 years. The
material herein builds on the World Bank Technical Note “Reform of Polish Covered Bonds” produced in
2009 (with the associated memorandum of 2010), as well as on a broader World Bank Polish Housing
Finance Policy Note of 2006. For avoidance of repetition, messages from those documents are used in
this Note directly, as appropriate.
At the same time, the discussions with and the commentary from the Polish authorities during and after
the Mission showed that certain aspects of the MCB framework and broader housing finance market seem
to merit additional clarification. Therefore, several key concepts are addressed with more detail, e.g. asset
encumbrance, prudent mortgage underwriting policy and practice, MCB issuer models, and comparison
of MCB to Residential Mortgage Backed Securities (RMBS).
The Polish authorities by virtue of proximity to a number of relatively advanced jurisdictions have access
to the wealth of mortgage finance information; they are encouraged to take advantage of this and consider
consultations with the EU and CIS market stakeholders, including regulators and trade groups. As the
policy and practice dialogue on the mortgage principles and practices, including micro and macro
regulation has been very active in recent years, direct and ongoing contact may be valuable in
crystallizing the development and stability agenda for Poland.
This Note primarily addresses the mortgage market development objectives, although the regulator is
encouraged to consider the use of the macro prudential tools at its disposal for institutional and systemic
stability of the mortgage sector as the current portfolio outstanding is large and risky and thus presents a
negative performance outlook. The rest of the material is structured as follows. Firstly, brief background
information on recent Polish mortgage sector evolution provides context. Secondly, details of certain
regulatory initiatives are discussed with the view on potential strengthening. Thirdly, suggestions for
modernization of certain features of the current capital market funding framework are provided;
specifically on expanding the issuance of MCBs to the universal banks as well as bringing the framework
in line with the latest global best practices.
1 The team was led by Luc Everaert (IMF) and Brett Coleman (World Bank) and included Karl Driessen, Nancy Rawlings,
Yinqiu Lu, Jorge Chan-Lau, Rishi Ramchand (all IMF), Katia D’Hulster, Heinz Rudolph, Andrey Milyutin, John Pollner, Ignacio
Tirado (all World Bank), as well as external experts David Walker (Canadian Deposit Insurance Corporation), Monnie Biety
(independent consultant), and Fernando Montes-Negret (former World Bank and IMF). This Note is prepared by Andrey
Milyutin (World Bank, FCMNB).
5
Executive Summary
Poland’s financial system broadly, and the mortgage market in particular, have so far demonstrated
resilience to the ongoing global financial crisis and the Euro Area turmoil. Vulnerabilities lie in the
exposure to foreign exchange (FX) risk and foreign investors, which in turn may pressure bank funding,
especially in the background of absent long-term domestic funding sources, negative cyclical outlook and
declining residential real estate prices. Asset quality has become a higher priority on the supervisory
agenda to address persistent non-performing loans and cyclical deterioration in credit quality. This has a
particular importance for the large and risky mortgage portfolio outstanding; furthermore, a recent
loosening of underwriting standards could contribute to rising mortgage NPLs and thus to institutional
and systemic risks.
Establishing a robust capital market funding framework including MCBs and RMBSs can address the
funding needs and mitigate the FX and asset quality risks facing the banking system. A number of recent
regulatory initiatives strive to remedy the situation, focusing on improved asset quality, loan
transferability, and reviving the specialized mortgage banks. A new large-scale housing subsidy program
is designed to improve targeting and efficiency while potentially mitigating fiscal risks.
The Polish regulators are advised to adopt a comprehensive approach towards further development and
stability of the mortgage sector. Firstly, the risks of the current portfolio may need to be aggressively
addressed in close cooperation with the lender community. The combination of multiple risk factors, e.g.,
FX and adjustable rate mortgages (ARMs), inefficient loan transfer and foreclosure procedures, a 20 %
PLN portfolio share of subsidized loans, low PLN loan seasoning add to the risk profile of the mortgage
portfolio. In the context of high and rising unemployment and prospects of an economic slowdown, this
presents a negative portfolio performance outlook.
Secondly, strengthening the mortgage lending regulatory framework is also advised, primarily in the area
of prudent underwriting and servicing. As the development of the proper capital market system is likely a
lengthy process, alignment of the Recommendation S with the current best global practices may be
necessary to avoid the repeat of the risky practices of the past and to support sustainable growth of the
mortgage portfolio.
Third, on a forward-looking basis, the Polish regulators are advised to consider that the high risk profile
of the current portfolio may in large part be due to the absent capital market funding channels. Thus, the
development of a modern framework of the MCB and RMBS instruments may contribute to systemic and
institutional development and stability. In this regard, congruence with the current global legal, regulatory
and policy trends is advised.
Lastly, the Polish market stakeholders are encouraged to actively seek information exchange and
cooperation with their regional counterparts, particularly in the countries where the mortgage markets
have experienced significant negative events in the context of the ongoing financial crisis.
6
Key Recommendations
Recommendation Notes
SHORT TERM
Address the risks of
the mortgage
portfolio outstanding
KNF and NBP, in close coordination with the lenders, are advised to adopt an
aggressive proactive approach to the credit risks of the mortgage loan stock.
Specifically:
1. Conduct a fact-finding exercise to obtain granular loan-level mortgage
portfolio data to identify the components of the portfolio (both per lender and
macro) where the risks are the most apparent –based on past experience and
on plausible economic scenarios;
2. Design appropriate watchlist procedures which would apply either on a
per-lender basis or per portfolio component. Certain prudential adjustments
may be needed, as appropriate;
3. Incentivize the lenders to conduct loan modification programs, if such
are found necessary and feasible. Address the public awareness concerns and
ensure appropriate consumer protection.
SHORT TERM
Strengthen
regulatory
environment for
mortgage lending
KNF are advised to consider fine tuning the current draft of Recommendation
S. Specifically:
1. Risk factors should not be allowed to stack, e.g., for a given loan,
maximum values of both DTI and LTV should not be allowed;
2. The decision to exclude DTI limits from the Recommendation should be
reconsidered;
3. The use of credit risk insurance in connection with LTV limits needs to
be clarified and stringent conditions applied to such insurance vis-à-vis related
party transactions, pricing, and other regulatory treatment (e.g., capital
adequacy ratio or provisioning rules applicable to such insurance).
MEDIUM TERM
Establish a modern
and robust capital
market funding
environment as
regards to mortgage
finance.
Market stakeholders are advised to adopt a comprehensive approach to
developing a modern capital market funding channel for mortgage finance.
Specifically:
1. Modernize the enabling environment to eliminate the legal and tax
obstacles of mortgage rights transfer;
2. Modernize the MCB framework to allow the universal banks, along
with the specialized banks, to issue MCBs, and include the latest features on
prudent and efficient treatment of such challenges as asset encumbrance,
potential fiscal liabilities, prudent loan eligibility criteria, robust cover asset
monitor functionality, issuer licensing, and MCB integration in the bank
insolvency framework;
3. Establish a modern securitization framework. The draft securitization
law will benefit from congruence with the ongoing EU and US regulatory
initiatives aimed at strengthening the quality of mortgage loans and the
securitization instruments (e.g., FSB guidance, Dodd-Frank Law and QRM,
similar EU initiatives).
7
Mortgage Market Context
1. The Polish mortgage market has so far been spared the negative effects of the global financial
crisis, yet significant risk factors and inefficiencies prevail. The portfolio outstanding is just over
EUR 80 Billion (approximately 1.6 million loans), representing 32 percent of the banking loan book. The
portfolio is the largest in Europe and Central Asia (ECA) both in absolute terms and in the share of the
banking book which is double the regional average and on par with CEE-5 (Czech Rep., Poland,
Hungary, Slovakia, and Slovenia). The market is significantly larger than in neighboring countries –
Russia (EUR 40B), Hungary (EUR 23B), Czech Republic (EUR 20B), Slovakia (EUR 12B). In the
broader EU context the market size is comparable to Austria (EUR 83B), Finland (EUR 81B) and Greece
(EUR 78B).2
2. More than half of the portfolio is in FX (primarily CHF), although FX originations have been
negligible since 2010 (Figure 1). In the absence of long-term domestic financing options, mortgages are
funded by deposits or interbank loans (mainly intragroup loans from foreign banks to Polish subsidiaries).
Mortgages typically have adjustable interest rates which reset every 1 to 6 months (set as the interbank
rate plus a margin fixed for the lifetime of the loan) and tenor of up to 30 years. This mortgage product
profile is typical for many ECA countries, where in the early 2000s the large EU banking groups
established significant presence in the sector and thus promulgated EUR or CHF ARM products. At the
same time Polish market stakeholders have been attentive to the FX risks in the aftermath of the acute
financial crisis phase in 2009, and FX originations have been drastically reduced since.
3. The average portfolio loan-to-value ratio (LTV) is over 80% (95% for FX loans), the share of
mortgages with LTV > 100% is high at 27% by volume, over 38% of the volume of FX loans have LTV
over 100%, 19% of FX loans have LTV over 130% and 10% of FX loans have LTV over 150%.3 FX
loans constitute 56% of the aggregate portfolio. These are conservative data as LTVs were captured at
origination and the housing prices have significantly declined since 2007-2009 when a large share of the
FX loans was originated. Thus the LTVs may need to be actualized in order to present accurate
2 Source – Hypostat, 2012 figures, 3 Source – KNF, 2013 for portfolio and 2012 for LTV composition.
Figure 1. Mortgage Portfolio Dynamics [PLN MM] Figure 2. Mortgage NPL Dynamics [%]
Source: KNF
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
FX mortgages PLN mortgages
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
4.5%
Weighted average PLN mortgages
FX mortgages
8
information both to the lenders and the regulators. The intersections of the risk factors present particular
concerns, e.g., 27% of the portfolio with LTV>100% and FX ARM mortgages. Significant further
analysis of the granular data is required, as discussed below.
4. While weighted average NPLs are relatively low at 2.6 % (4 % for PLN loans), there is a clear
upward trend (Figure 2). In the global context the current Polish NPL levels can be seen as rather benign,
especially when compared to the countries that are experiencing significant negative effects of the global
financial crisis, e.g., Hungary, Ireland, Spain, US, etc. 4 However, the combination of multiple risk factors
(e.g., high LTV, FX and ARMs, inefficient loan transfer and foreclosure procedures, a 20% share of the
PLN portfolio of subsidized loans and low seasoning) add to the portfolio risk profile. In the context of
high and rising unemployment and prospects of an economic slowdown, this presents a negative
performance outlook.
5. Domestic long-term funding sources are absent; the system of the specialized mortgage banks has
become irrelevant. The two remaining specialized mortgage banks have almost entirely switched to the
commercial mortgage sector. Universal banks do not have access either to the MCB or the securitization
funding channels and instead rely on short-term deposit liquidity for mortgage lending. In addition to
creating maturity gaps in the banking system, the absence of a modern mortgage financing mechanism
has contributed to the mortgage portfolio risks. Since the terms and conditions of mortgages from the
universal banks can be more flexible, mortgage banks have little chance of sizeable origination. The
transfer of mortgage rights between creditors is inefficient due to issues in taxation, re-registration,
nontransferability of the Banking Execution Title, and the requirement to obtain borrower consent to the
transfer. The mortgage banks thus do not have the sufficient volumes of the conservative MCB-compliant
mortgages due to challenges in origination or purchasing and therefore cannot issue large and predictable
volumes of covered bonds. The universal banks have no incentive to originate MCB-compliant loans
(since they cannot issue MCBs and the transfer of mortgage loans is inefficient) and thus have the
flexibility to relax their underwriting criteria.
6. There is an active policy and regulation dialogue among market stakeholders, and a number of
recent regulatory initiatives are aimed at strengthening the sector. The authorities seem to continue to
expect the existing system of mortgage banks to succeed notwithstanding long domestic experience to the
contrary and the evident global and regional trend of modernization. Three working groups have
produced legal and regulatory drafts aimed in large part at preserving the current system of specialized
mortgage banks, e.g., a scheme to allow universal lenders to use the balance sheet of their mortgage
subsidiaries and thus to avoid the re-registration costs. However, further revisions in several areas are
suggested.
7. The draft Recommendation S is a step forward with the introduction of explicit LTV limits and
the requirement to match the currency of the borrower’s income with the currency of the mortgage.5
Certain taxation inefficiencies of loan sales to the mortgage banks have been recently resolved, and the
electronic centralized re-registration mechanism is expected to be available in 2013.
4 NPLs in US - 6.8%, Serbia - 7.35%, Ireland -11.9%, Russia – 2.1%, Turkey – 2%, Hungary-15%. Sources – MBAA, respective
Central Banks, AHML. Latest available data used (12/2012)
5 Subsequent to the mission, the FSAP team was informed that the amended Recommendation S was published in
June 2013.
9
8. Large-scale housing subsidy programs have been recently redesigned to improve efficiency. The
2005-2012 Rodzina na Swoim scheme was large with 182,000 loans (PLN 35B) extended (estimated 20%
of the total and 30% of PLN originations by number of loans or 20% of the current PLN portfolio
outstanding). However, its 8 year 50% interest rate subsidy for ARM loans has created an uncertain future
fiscal liability. The yet to become official next subsidy program (MDM) seems to be an improvement with
strengthened key features – subsidy towards the downpayment instead of the interest rate, better real
estate price targeting, and focus on young families. The expected volume of MDM is estimated at
115,000 loans but may have a significant impact on the market as the overall originations decline. The
authorities may wish to consider establishing prudent origination and servicing practices for MDM-
eligible loans for market development and stability reasons.
Addressing the Risks of the Current Mortgage Portfolio
9. In analyzing the existing mortgage portfolio, it appears that a significant share was originated
under relaxed underwriting policies and practices, which was made possible in part due to availability of
funds from the parent EU banking groups or retail deposits and the banks’ race for the market share as
well as to the bottom of the quality. In the context of large individual and aggregate mortgage portfolio
with risky characteristics of high LTV, FX, tracker
ARMs – overlaid with declining property prices and
rising unemployment - the market stakeholders are
advised to consider robust and aggressive portfolio
management activities.
10. The authorities and the banks together should
consider a proactive, forward looking approach to
dealing with delinquent mortgage loans (both stock
and future originations) up to and including
foreclosure and borrower eviction. The special
servicing policies and practices6 need to be addressed
from the market stability and further development
perspectives.
11. The important balance to strike is between predictability of the special servicing process – both
under normal, as well as emergency circumstances – and the strength of the collateralization of the real
estate which is the cornerstone of mortgage lending. This balance has both institutional and macro policy
implications, as at stress times
imprudent foreclosure actions
by a number of individual
institutions quickly spread
among the industry and impact
the housing markets. Note the
examples of Atlanta, GA when
6 Special servicing in the context of mortgage finance denotes policies and procedures that lenders or their agents follow when
dealing with delinquent and defaulted loans. Specifically, such practices include contact with the delinquent borrowers, loan
modifications, legal proceedings, property sales, conversion to rental, etc. The canonical objective of special servicing is
maximization on the NPV basis returns to the lenders (minimization of losses).
Mortgage Portfolio – Key Ratios, December 2012
Average portfolio LTV 80%
FX portfolio average LTV 95%
Share of portfolio with LTV > 100% 27%
Share of FX loans with LTV > 130% 19%
PLN delinquencies 4%
FX delinquencies 1.8%
What is 1% of NPLs?
In the Polish context 1% increase in mortgage NPLs means that around
15,000 families are in default and are facing eviction.
In certain circumstances, e.g. spatial and institutional concentration, this can
present a significant social and institutional stability concerns.
10
in 2010 over 66% of the housing supply was in foreclosed properties put for sale by the lenders, or of the
robo-signing scandal in the US among virtually all major mortgage lenders and servicers.7
12. As mentioned before, Poland has a large – EUR 80B – stock of mortgages, many of which likely
were underwritten to poor risk management practices and policies. This notion is supported even by the
limited available aggregate data stratified only by LTV and currency. Even if significant NPLs have not
materialized yet, the regulators are well advised to proactively approach the issue of the risks of such
portfolio and not to wait until the industry shows high and unsustainable delinquencies as is the case in,
for example, Kazakhstan, Hungary, Ireland or Ukraine.
13. Additionally, the authorities are encouraged to consider the current unfavorable NPL trend and
composition. Specifically, PLN delinquencies are high at 4%, which is a significantly elevated level per
se, but especially so in the context of the unseasoned8 portfolio with a 20% share of subsidized loans. The
NPL share of this portfolio was growing almost linearly since 2009 – at the speed of 0.1% per month,
which is troublesome given the large increase in PLN originations since 2010, which under normal
circumstances would have arithmetically reduced the reported delinquencies.
14. A suggested approach would be to scale the micro portfolio management and watchlist practices
to the macro, country level and codify some of the better techniques into policy. Note that watchlist
procedures apply to the loans that are current, but due to a combination of the present or expected
borrower and loan characteristics may present an elevated risk profile. Specifically, the actions to take
are:
a. Conduct ex ante facto portfolio analysis. The regulator should request appropriately
stratified mortgage portfolio data from all of the significant lenders [80-90% of the
aggregate portfolio] with a particular focus on the lenders with the largest relative
exposures. The grouping should clearly show components of the portfolio with high
potential risks, especially in the layering of risk factors, e.g. FX, LTV in excess of 100%,
properties located in economically depressed areas or in areas with significant House
Price Appreciation (HPA) declines, borrowers from the industry sectors that have or are
likely to experience significant unemployment increases, etc. Additionally, lenders
should be consulted on their risk management and analytical observations as regards to
credit risk drivers. In the Polish context, actualization of the LTVs and DTIs may be
warranted for certain portfolio strata. A formal Basel 2 compliance expected loss model
should optimally be constructed.
b. Design of the appropriate risk management measures. Depending on the findings of the
previous step – volumes, locations, number of lenders and borrowers affected, etc. – the
regulator and the lender community may consider jointly designing an appropriate loan
modification action plan. This plan could include standardized modification programs
and the required regulatory amendments – likely temporary.
7 See for example www.housingwire.com/news/2010/10/08/robo-signer-effect-housing-market-reaching-critical-mass and
business.time.com/2010/10/19/will-bankers-go-to-jail-for-foreclosure-gate/?xid=rss-topstories 8 In mortgage terminology a loan is considered “seasoned” after 12 months. A frequently used rule of thumb is that mortgage
defaults rise during first 3-5 years of loan life and plateau thereafter.
11
c. Conversion of the loan (modification) to less risky and more prudent mortgage product
may take a number of formats.9 For example, in the current environment of low interest
rates, the lenders may be encouraged to extend the ARM reset periods from one month or
to implement smoothed mechanisms of the interest rate calculation. Another example
would be conversion from FX to PLN or from ARM to FRM. A typical loan modification
measure is lengthening of the loan term and re-actualization of the LTV via property
appraisal.10 In any case, modification measures may have options appropriate to different
borrower strata and should not result in an absolute increase in the mortgage payment.
d. However, regulatory requirements, for example, to lengthen the ARM reset terms, may in
part serve the same purpose. A significant ex-ante impact analysis should be done, and
measures should be taken to ensure gradual availability of PLN funding sources. Note
that the measures may be institutionally specific, e.g., if a particular lender is found to
have a large share of particularly risky mortgages with spatial concentration. They may
also be systemic, e.g., a large share of the aggregate portfolio across many lenders is
found to have risky characteristics. In the latter case the regulator may consider providing
for appropriately beneficial treatment (likely temporary) of the modified loans in terms of
capital charges, provisioning or liquidity ratios.
e. Implementation. If the findings of the previous steps warrant a relatively universal
modification program – likely voluntary for borrowers – the selected procedures need to
be made public with an appropriate consumer awareness campaign; frequently the
regulator’s participation in such activities raises the perceived level of trustworthiness in
the eyes of the public and thus may increase effectiveness. Additionally, regulatory
oversight is essential to prevent prohibitively high lender fees and charges.
f. Care should be observed to avoid in appearance or substance any measures that would
lead to the mortgage portfolios being nationalized, as borrowers may fall under the
impression, frequently justified, that the State would not foreclose and evict in cases of
defaults. Additionally, banking policies and regulation as regards to such a program
should be clear and transparent to avoid potential moral hazard and adverse selection, i.e.,
eligibility criteria verification, transparent procedures, defined volumes and timetable,
seasoning of the loans, etc.
15. At the very least, even if the data show that no universal program is appropriate and thus no
policy measures are required, the authorities would gain a critically important understanding of the
stratified composition and performance of the national portfolio. Besides maintaining this awareness on
an ongoing basis by periodically [optimally monthly, likely quarterly initially] collecting the loan level
data, the regulator is encouraged to analyze it in the context of the Polish economic performance, e.g.,
overlaying it with the location-specific HPA data to understand true LTV situation both systemically and
institutionally.
9 Examples of such programs are in Hungary, Ireland, Russia, and US where they were either codified by regulation/law or
industry-driven. 10 Re-calculation of the LTV in the current environment of lower property prices compared to 2007-2008 may be beneficial for
lenders in case the regulator adopts capital allocation requirements in relation to the quality of the mortgage loans similar to the
ones discussed in paragraph #27.
12
16. From the policy and regulatory perspective, it is possible to incentivize the lenders to pursue
robust watchlist and modification practices by certain capital allocation and provisioning measures,
consumer protection and disclosure requirements, and aggressive compliance measures. It is critical for
the regulator to maintain detailed and complete HPA awareness as well – overlaid with the aggregate
mortgage portfolio evolutions to be in a position to assess the situation and react with targeted policy
measures as appropriate.
Why do NPL levels fall?
The Polish regulators are encouraged to monitor the NPL situation beyond the aggregate headline number and
remember that delinquency ratio evolutions may be in large part due to the simple arithmetic “tricks” related to
the changes in origination dynamics. Only granular stratified ongoing portfolio surveillance allows for
appropriate design of the focused and determined lender activities in loan modification and workout, which
ultimately may bring about improvements in the portfolio risk profile.
13
Strengthening the Mortgage Regulatory Environment
18. The Polish regulators are encouraged to consider that the risky portfolio outstanding was made
possible in part by the absence of a viable capital market funding channel, which led to the situation
where regulatory guidance on prudent mortgage lending and servicing was the only mechanism to ensure
quality underwriting. Globally, however, such reliance on regulation has proven to be insufficient, as even
the largest mortgage markets have discovered the influence of often perverse and asymmetric lender
incentives, which led to product aberrations, such as extreme cases of “Alt-A” in the US or exotic FX
(including JPY) ARM mortgages in ECA.
19. One of the major lessons from the ongoing
financial crisis, ignited in part by the systemic
failures of the US sub-prime mortgage market, is that
ill-conceived mortgage products and weak
underwriting and servicing can affect entire financial
systems even if such practices may be initially
limited to specific institutions. In the broad trend of
revising and strengthening the “basics” of housing
finance – mortgage loans as such – increased
attention to underwriting practices has been evident
among market stakeholders and regulators. The need
to define and implement adequate underwriting
standards is of particular importance in the ECA
region where funding constraints or incomplete
mortgage market infrastructure may conflict with the
soundness of lending.
20. The market-wide consequences of poorly
conceived and implemented origination policies and
practices, out of proportion with the micro level of
the triggering factors, are costly in terms of systemic stability and institutional bailout. Additionally, they
halt the deepening of housing finance and demonstrate the importance of sound and prudent lending
standards for the sustainability of market development. Since 2007, recommendations and regulatory
adjustments related to strengthening of mortgage lending framework have been issued in many countries,
including the US, EU,11
the UK,12
and Hungary,13
as well as internationally, e.g., by FSB and BIS. They
provide an updated foundation for healthy policies, which however needs to be customized to the
specificities and development level of each national market. Some the most critical principles are:
21. Assessing borrowers’ ability to repay is the primary consideration when lending for housing.
Elements of this include:
a. gaining accurate knowledge of the borrower’s income – both volume and type;
b. taking all the existing borrower’s obligations into account, and
11 EU Consumer Credit Directive 2008/48/EC 12 Mortgage Market Review: Responsible Lending (FSA, 2010) Mortgages: Conduct of Business (FSA, 2007), Handbook for
Mortgage and Home Reversion Brokers (FSA, 2008) 13 Consumer Credit Act 162/2009 and Decree 361/2009 (12/2009)
… ensure that lenders consider more conservative
underwriting criteria to compensate for situations where
the underlying risks are higher.
For example, more conservative underwriting standards
(e.g. LTV ratios or servicing requirements) could be
considered where:
…there are considerable risks that an asset price bubble is
building up in the property market as a whole or in specific
segments or geographical areas;
…the loan is in a market segment that, compared with
other mortgage loans in that jurisdiction, tends to perform
worse than average in a property downturn (depending on
the jurisdiction, examples of such a market segment might
include luxury apartments, buy-to-let investors, second
homes, cash-out refinancers, etc.)…
…Jurisdictions may want to impose absolute minimum
levels of particular dimensions of mortgage underwriting
standards below which no mortgage would be deemed
acceptable, irrespective of the settings across the other
dimensions.
Financial Stability Board Mortgage Underwriting
Principles (2012)
14
c. in the case of ARM or FX mortgages - assessing the future repayment capacity based on
conservative assumptions and periodic stress tests. Understandably, enforcement of
forward-looking positive covenants, e.g., related to minimum DTI levels, raises
challenging loan servicing issues and should be best approached by lenders on an
individual basis vis-à-vis specific borrower circumstances.
22. Protection against price cycles particularly from price bubbles. LTVs should be set at levels that
do not reflect the extrapolation of an appreciation trend in the future, but instead should reflect realistic
assumptions of recovery rates. Some jurisdictions have established regulatory “hard” limits, an approach
that depends on the specificities of a particular market.
23. Prudent origination must rely on certain supporting infrastructure, such as:
a. The availability of credit registers, including both negative and positive information and
utilization of information from the bureau during underwriting and servicing processes;
b. Reliable, standardized and independent appraisal capacities to ensure the accuracy of
LTV values – including typically on-site appraisal at loan origination and periodic desk-
top portfolio reviews;
c. Credible and time-predictable foreclosure process;
d. Availability and utilization of appropriate insurance products and mechanisms, including
coverage for lender credit risk, property hazards and borrower health.
24. Compensating for, rather than adding risk factors. Particular attention should be exercised by
regulators and supervisors to “risk layering” or adding several risk factors within the same transaction,
which is one of the most damaging practices associated with the US sub-prime market. Broadly, lenders
should establish a certain normalized “prime” level of risks associated with their mortgage portfolio –
both in terms of borrower profile and loan terms and conditions. Variations of the products should strive
to maintain a symmetrical approach to modifying such level, e.g., when a riskier borrower strata is
targeted, loan features should offset such increased risk, and not simply, as is common practice, add more
credit risk by increasing the interest rate. Compensatory measures may include modifying underwriting
criteria, enhancing loan servicing, avoiding FX, hybrid and ARM features, requiring additional or
modified insurance coverage, etc. Examples of such asymmetric combinations, which need to be avoided,
include:
a. ARM loans to borrowers with irregular incomes or in combination with high LTV and
DTI;
b. gimmicks used to make debt affordable to lower income borrowers but that often have a
delayed, time bomb impact of their solvency such as bullet repayments, repayment
profile involving negative amortization, or initial teasers rates; and
c. FX loans in combination with high LTVs and DTI.
25. Draft Recommendation S (as per the version made available to the Mission) is a significant step
forward and incorporates much of the global guidance on prudent mortgage underwriting. However,
certain aspects of it should be fine-tuned. In particular, the KNF inclination to leave the policy and
practice matters of the decisions related to DTI to the lenders, and the currently vague description of the
mortgage insurance role in underwriting are advised to be reconsidered.
15
26. Firstly, the decision to exclude DTI limits from the Recommendation leaving the DTI limit and
guidelines to individual lenders should be reconsidered as it may lead to risky and undesirable
consequences, such as:
a. There is a risk that multiple incomparable DTIs in the system will emerge, as different
lenders may use different definitions and calculation methods. A critical aspect of this
risk is that this metric may become irrelevant both in the aggregate (affecting macro
policy actions) and comparatively between institutions, leading to difficulties in portfolio
pricing, prudential oversight or secondary mortgage transactions.
b. Delegation of the DTI-related policy to the lenders’ Boards may create a conflict where
KNF, in case it questions a given bank’s DTI practices, has no legally enforceable powers
to deal with the bank’s Boards;
c. DTI and LTV are interrelated metrics of the credit risk of a mortgage loan; avoiding
guidance on one of them creates challenges for understanding the overall credit risk
profile of the micro and macro portfolio. This in particular is important for the regulator
in a down cycle, as both the countercyclical steps and emergency asset quality related
measures may need to be taken rather swiftly.
d. As DTI is highly susceptible to borrower and lender manipulation both at origination and
during the life of a loan, avoiding regulatory guidance on DTI leaves ample room for
risky practices and thus creates challenges for understanding the overall credit risk profile
of the micro and macro portfolio.
27. Additionally, mortgage loan risk factors should not be allowed to stack, e.g., for a given loan,
maximum values of both DTI and LTV should not be allowed. KNF is encouraged to consider that the
interconnectedness between several key parameters of a mortgage loan – as regards to the credit risk as
well as macroeconomic impact – has been established and used by regulators and the industry in other
jurisdictions. Below are a few examples that may illustrate this concept.
a. The canonical definition of a credit risk as a product of the probability of a negative
credit event and the severity of such event, in the mortgage context traditionally uses
LTV and DTI as proxies for the loss severity and event probability, respectively. In other
words, while the LTV ratio traditionally serves as an indicator for the potential loss of the
lender in case of a borrower default, the DTI is seen as an indicator of the borrower’s
ability to pay, which, together with the borrower’s incentive to pay (influenced by such
factors as property price evolutions, LTV, family circumstances, occupation of the
collateralized property, etc.) drives the probability of a default event.
b. LTV and DTI ratios are highly correlated, i.e., in case of an economic downturn,
typically real estate asset prices fall (LTV increases) and unemployment rises, reducing
the borrowers’ incomes (DTI increases).
c. Regulators in the EU have recognized the importance of stacking or “layering” of the
various risk factors in mortgage loans, evidenced in part by the FSB and BIS, and
illustrated by a number of related reports and research.14
14 2011 Liikanen Report http://ec.europa.eu/internal_market/bank/docs/high-level_expert_group/report_en.pdf, 2010 BIS The
Joint Forum Review of the Differentiated Nature and Scope of Financial Regulation Key Issues and Recommendations
16
d. Capital market funding intermediaries in the US have recognized the importance of a
matrix approach to mortgage loan terms and conditions. For example, Federal National
Mortgage Association (FNMA) explicitly defines the procedure for DTI calculation and
uses the matrix of DTI, LTV and credit score to determine borrower eligibility.
e. Mortgage insurers in the US, e.g., MGIC, also use a matrix approach to borrower
eligibility combining loan amount, DTI, LTV and the credit score.
f. Capital market funding intermediaries in Russia have also recognized the interrelated
nature of the various loan terms and conditions, e.g., the Agency for Home Mortgage
Lending (AHML) in the Federal Standards for certain products uses the matrix of the
loan amount, LTV, location of the property, and a number of other factors.
28. At the very least there should be a standard methodology for calculation of this ratio as well as a
prudent upper limit [40-45 %]. At the same time, it is clear that lenders and investors would have appetite
for mortgages with a different credit risk profile and would want to have flexibility in designing products
which would fit in the loan parameter matrix. Thus DTI regulatory guidance may take the form of capital
allocation requirements which would be dependent either on the nominal values of this parameter or
values outside of a certain prudent range. Additionally KNF are advised to consider that in general DTI,
as an aspect of the borrower’s eligibility, is a relatively “softer” parameter when compared, for example,
to LTV. This softness is due to significant uncertainty as to current and future borrower income. Thus the
regulatory and policy guidance on the DTI limits and the methodology for income assessment (e.g.,
approaches for including multiple income types and projecting future borrower income) need to be
carefully designed, implemented, and enforced – preferably in cooperation with the major lenders in order
to establish a statistically meaningful basis for risk assessment and thus capital consumption guidelines.
29. Additionally, the Polish regulators are encouraged to consider that the true meaning of the
prudent loan parameters and their efficacy as macro policy tools depends on the prevailing mortgage
products. For illustration, see the table below.
Mortgage Loan Type Stylized forward looking DTI scenario Stylized Corollary
Fixed rate, local
currency fully
amortizing mortgage,
e.g. US, Russia
The periodic mortgage payment is fixed and known
with certainty for the life of the loan. In real terms
declines with inflation. Incomes broadly follow
inflation, lag possible. Prevailing interest rates do not
affect payment amount. Ceteris paribus DTIs decrease.
Credit risk potentially less sensitive to
elevated initial DTI parameters in certain
cases of expected income increase, e.g.,
young family, government employees,
etc.
Local currency loans
with multiyear interest
rate reset periods, e.g.,
Canada, France,
Germany
The periodic mortgage payment is fixed and known
with certainty for 3-5 years. In real terms declines with
inflation during reset period. Incomes broadly follow
inflation, lag possible. Prevailing interest rates do not
affect payment amount during rest period. At reset
time, prevailing interest rates used, borrower and
property underwriting may be actualized. Ceteris
paribus DTIs likely level with reduced cyclical
sensitivity.
Credit risk potentially less sensitive to
elevated initial DTI parameters in certain
cases of expected income increase, e.g.,
young family, government employees,
etc. Actualization of underwriting a
significant risk management factor.
http://www.bis.org/publ/joint24.pdf, 2012 FSB Principles for Sound Residential Mortgage Underwriting Practices
http://www.financialstabilityboard.org/publications/r_120418.pdf
17
Monthly tracker
adjustable FX mortgage,
e.g., Poland, many ECA
countries
The periodic mortgage payment is fixed and known
with certainty for 1 month. Sensitivity to FX and
interest rate volatility beyond lender discretion.
Incomes broadly do not adjust monthly. At reset time
borrower and property underwriting is not actualized.
Ceteris paribus DTIs likely erratic and sensitive to
factors beyond lender or borrower control.
Credit risk potentially very sensitive to
DTI due to direct translation of FX,
interest rate and cyclical volatility.
Lender discretion in rate resets may be
risk mitigating measure, e.g. BY, JO, RU.
Negative amortization may happen in
some product designs.
30. In reviewing the global experience with DTI limits, the following qualifications should be noted.
Firstly, in many cases the marketplace – often related to the capital market funding – imposes
underwriting guidance on lenders, including for DTI. Secondly, as discussed above, in many instances
DTI limits are connected to the other loan parameters, and thus should not be viewed as a stand-alone
factor. Note that in cases of either regulatory or market-driven DTI limits, the relevant parties provide
exhaustive procedures for calculating the ratio itself, e.g., what types of income to consider, what
documentation to use, etc. Thus, subject to these qualifications, the following DTI limits illustrate broad
tendencies:
US 2010 Dodd-Frank Wall Street Reform and
Consumer Protection Act, 2013 Regulation Z15
Regulation
back end DTI of 43% subject to certain qualifications and
conditions, matrix approach
US (FNMA)16
Market
intermediary
back end DTI of 45% subject to certain qualifications and
conditions, matrix approach
US (FHA, MGIC)17
Mortgage
Insurers
back end DTI of 43% subject to certain qualifications and
conditions, matrix approach
Russia (AHML)18
Market
intermediary
back end DTI of 45% subject to certain qualifications and
conditions, matrix approach
US VA Loans Subsidy
program
back end DTI of 41% subject to certain qualifications and
conditions
31. At the same time, the Polish authorities are encouraged to view the best global practices as
references only and aim to take further steps, specifically as regards to periodic DTI and LTV
actualization and usage of such data both for risk management and macroeconomic purposes.
Specifically, the FSB guidance which explicitly instructs regulators to adopt policies and practices which
prevent risk layering should be of the most relevant to Poland as a member of the EU.
32. Third, the brief mention in the draft Recommendation S of the use of credit risk insurance in
connection with LTV limits needs to be clarified and stringent conditions applied to such insurance vis-à-
vis related party transactions, pricing, and other regulatory treatment (e.g., capital adequacy ratio or
provisioning rules applicable to such insurance). Global market and regulatory practices in this regard
vary, although certain principles may be used as a basis for consideration by the authorities:
a. Mortgage Insurance (MI) is a catastrophic type insurance and is often sold by monoline
insurers; re-insurance opportunities are rather limited globally, which puts significant
capital and balance sheet pressures on such monoliners;
15 www.gpo.gov/fdsys/pkg/PLAW-111publ203/html/PLAW-111publ203.htm www.consumerfinance.gov/regulations/ability-to-
repay-and-qualified-mortgage-standards-under-the-truth-in-lending-act-regulation-z/ 16 www.fanniemae.com/content/guide/sel011713.pdf 17 www.fha.com/fha_requirements_debt 18 www.ahml.ru/ru/borrower/ipProg/standart/
18
b. MI may take the form of a guarantee (i.e., immediate payment continuity protection for
the lenders), loan-level insurance (i.e. compensation of loss to the lender after the default
resolution), and pool-level insurance in cases of capital market transactions;
c. Loan-level MI is frequently seen by the authorities as a beneficial socially oriented
measure and thus risk is frequently mispriced, creating additional institutional stability
and sustainability challenges which transform at the times of distress to the systemic
stability issues and fiscal liabilities;
d. Prudential treatment of MI coverage needs to be carefully considered, as to its quality,
target market segments, presence of a subsidy, quality of the underlying mortgages and
lender incentives. Especially in the context of Basel 2 and 3, MI coverage could be
treated equally with actual LTV coverage as regards to capital adequacy and
provisioning, subject to high quality of the risk transfer effected by the insurance product
and procedures;
e. Broadly, lenders globally seem to prefer increasing the risk premia, e.g., interest rates in
the case of high LTV loans as opposed to utilizing MI in the absence of regulatory or
market-based incentives.
33. The Polish authorities are strongly encouraged to consider the question of MI utilization
thoroughly, including in the context of the upcoming MDM subsidy program. Pricing and regulatory
treatment of this insurance type may be the key challenges and KNF together with MoF are strongly
encouraged to seek diverse international expertise on this subject. Countries with significant MI
experience in both public and private sectors include Australia, Canada, and US.
19
Modernization of the Mortgage Covered Bonds System
34. Polish authorities are encouraged to consider that in the absence of a capital market funding
channel the lenders have been free to pursue risky product development and origination policies and
practices limited by only the KNF regulatory guidance. In jurisdictions with developed capital market
funding, e.g., US, Canada, certain EU members, Russia, Korea, and Japan, the guidance from such
channels (legal, regulatory or practitioner-driven) in large part determines the terms and conditions of
mortgages.
35. It follows that when (and if) the large Polish mortgage lenders, which are universal banks, are
allowed to issue MCBs, the MCB eligibility criteria will define at least a portion of the originations. The
development of a securitization funding channel will also lead to additional underwriting guidance for
mortgages which are to be used in RMBS transactions. Based on a broad global experience, typically less
than 50% of the funding for mortgage lending comes from capital market channels, even in the highly
developed countries. Thus, some of the mortgages even in the environment of diversified funding sources
will still have terms and conditions defined largely by the prudential regulations.
36. In 2012 the Polish market stakeholders formed three working groups under the chairmanship of
KNF - on covered bonds, on securitization, and broadly on long term funding sources. The Mission was
advised by the participants of these groups that all three were successful in identifying key legal and
regulatory issues which prevent the development of a long term funding system in Poland – both broadly
and specifically to mortgages. In early 2013 all three groups delivered reports to the authorities listing the
issues and outlining plans to conduct further selection and deliberation and ultimately draft an action plan.
The main thrust of the reports is reportedly in reviving the currently nonfunctional system of the
specialized mortgage banks by addressing certain legal and regulatory deficiencies, e.g., a scheme to
allow universal lenders to use the balance sheet of their mortgage subsidiaries and avoid the re-
registration costs; finalization of the electronic centralized re-registration mechanism, elimination of VAT
on mortgage loan sales to the specialized banks, etc. The text of the reports was not made available to the
Mission, thus further commentary on the scope and substance of the proposed amendments is impossible.
37. The Polish market stakeholders enjoy a productive and rich exchange with the counterparts from
the EU jurisdictions with well-developed MCB markets; however, a number of specific MCB-related
questions seem to have been more important or less understood than others. Therefore, this Note will only
briefly cover the general MCB characteristics, focusing instead on the following questions:
a. Key MCB characteristics from a rating agency perspective;
b. Potential approaches to manage asset encumbrance;
c. MCB issuance models.
38. Ultimately, the Polish regulators are encouraged to pursue a comprehensive approach to
strengthening the quality of their mortgage lending sector. As discussed in the section above, modernizing
the Recommendation S is an important immediate element of such approach; development of a proper
capital market funding channel is another.
Key MCB Characteristics from a Rating Agency Perspective
39. Essential features of Covered Bonds from a rating perspective:
a. Credit rating of the issuer: the (implied) rating of the issuer forms the basis of the analysis
for the rating agencies and the floor for the MCB rating;
20
b. Strength of legal framework and/or contractual arrangement: issuing Covered Bonds
under a specific framework is preferred as, besides the legal certainty, this also gives the
raters comfort of regulatory support, supervision, and stability of the framework. In many
countries where MCBs initially were issued under general law (e.g., UK, The
Netherlands, and France), specific MCB laws were later implemented. On top of the law,
raters would like to see more details in the contracts on asset and liquidity tests (e.g., pre-
maturity test, extension periods), potential commingling, and set-off reserves etc.;
c. Recourse to a cover pool of high quality: MCB-investors need to have a preferential
claim on the cover pool in case of insolvency of the issuer of the MCBs. They will
require confirmation in a legal opinion. The only other creditors which can rank pari
passu at the same level as MCB investors are swap counterparties that hedge interest
and/or currency risk in the MCB program. Raters will impose a penalty in terms of
overcollateralization (OC) if other creditors rank at the same or higher level as MCB
holders. The cover pool is in general managed and serviced by the issuer/originator to
ensure assets in cover pool are eligible;
d. Supervision by the regulator and/or by the Cover Asset Monitor: Most cover pools are
monitored by a Cover Asset Monitor which is a trustee or qualified auditor and reports to
the regulator. The Cover Asset Monitor also checks that the asset tests are met over time.
40. Rating agencies in general have the following requirements to mitigate the risks embedded in
Covered Bonds:
a. Credit risk: the credit risk is assessed by rating agencies using similar techniques to
RMBS scoring models on a loan-by-loan basis (S&P and Fitch) or stratification table
basis (Moody’s). The credit risk needs to be covered by OC but in general it is not the
main driver of OC and accounts for less than 10% of the OC. In the MCB legislative
framework there are provisions for maximum LTV of mortgage loans eligible for
inclusion in cover pools - 80% (most countries) or even lower (60% Germany, 70%
Finland, 75% Norway and Sweden). Credit risk is also mitigated by the replacement of
the by the issuer. The cover pool should at all times meet the asset test with more
performing assets than outstanding Covered Bonds;
b. Liquidity risk: this is the main driver for the OC in MCB programs. Most Covered Bonds
have a fixed rate and a bullet maturity (often with a one-year extension period). On the
other hand, the cover pool consists of assets which amortize over time, pay a mixture of
interest rates, and have long maturities of up to 30 years. Therefore, the liquidity risk in
case of an issuer default is significant;
c. Market risk: interest and currency risk stemming from mismatches between value of the
collateral and Covered Bonds over time. This risk is in general mitigated by derivatives
or by providing more OC;
d. Counterparty risk: main one being issuer/originator: this is typically mitigated by relying
on the rating of the issuer or the parent supporting the issuer. The servicing risk in
general is mitigated by triggers that arrange for back up servicer activation once the
rating of the primary servicer falls below a certain level. For swaps and payment agents,
rating agencies apply their standard counterparty criteria, e.g., below a certain rating
(single A for AAA-rated CBs) counterparties need to be replaced, obtain a guarantee, or
21
provide collateral to prevent a downgrade of the MCBs. This is outlined in the standard
counterparty criteria report published and regularly updated by the raters. To hedge
commingling risk and set-off risk, raters will require reserves to be set up if the issuer’s
rating is below a certain minimum level;
e. Sovereign risk: the rating of the sovereign has emerged an in increasing number of MCB
programs as the limiting factor because the rating agencies cannot rate MCBs above the
sovereign rating ceiling. Moody’s allows 3 notches above the sovereign rating, Fitch four
and S&P up to six for Eurozone countries.
Approaches to Manage Asset Encumbrance
41. In a number of conversations during the FSAP mission, the Polish regulators broached the subject
of asset encumbrance. In their view, MCBs issued by universal banks may pose a stability and fiscal
liability risk since the high-quality assets in the MCB cover pool would be excluded from the issuer’s
bankruptcy estate and not available for resolution purposes. This situation, depending on the particular
circumstances of a given issuer, may lead to higher losses for the lower-ranked creditors and depositors of
the bank, the latter potentially increasing the fiscal exposure in case of the losses exceeding the capacity
of the deposit insurance guarantee.
42. The policy and regulatory dialogue as regards to covered bond asset encumbrance is developing,
and the recent surge in MCB issuance in the context of ongoing EU bank failures has further fuelled this
discussion. In this regard, the asset encumbrance discussion can be broadly approached from the
following perspectives. Firstly, the most straightforward mechanism to ensure that the issuing bank’s
depositors and creditors (besides MCB holders) are protected in case of bank insolvency is a hard limit on
the MCB cover pool relative to issuer assets. A number of countries have imposed limits on MCB volume
outstanding in order to better protect the interests of depositors, other bondholders and creditors. These
limits can explicitly be part of the covered bond legal framework, as for example in Australia,19
Canada,20
New Zealand,21
Italy,22
US23
; with the same issuance limit applied to all issuers. Alternatively, issuance
limits may be agreed on a case-by-case basis between the regulator and the issuer, e.g., in the Netherlands
and the UK. Note that the legal limit applies not to the MCB issuance volumes, but to the cover pool
assets, which are typically larger that the MCB volumes due to OC requirements.
43. The potential volume of MCB issuance by a given bank is not unlimited, as the availability of
high-quality mortgage loans (a subset of the overall mortgage origination) is a restricting factor putting a
cap on the actual issue volumes. The rating agencies have been recently raising their requirements on the
OC well above the legal limits which further reduces the available headroom for covered bond issuance.
Fitch’s 2012 study showed that more than 50% of covered bond issuers have a funding reliance (defined
as outstanding covered bonds as a percentage of total assets) of less than 10%. Only 1 in 5 issuers (almost
exclusively specialized mortgage banks) has a funding reliance of more than 20%. In addition to the cover
19 The value of the cover pool must not exceed 8% of the issuer’s assets for authorized deposit taking institutions (European
Covered Bond Council (ECBC)). 20 The maximum issuance limit is currently 4% of total assets (ECBC). 21 The value of the cover pool must not exceed 10% of total assets (ECBC). 22
Issuance limits depend on the capital strength of the issuing bank – though there’s no ceiling if total capital ratios are over 11
per cent and Tier 1 is above 7% (ECBC). 23
In the US there is a cover pool asset limit of 4 per cent of total issuer liabilities (Federal Deposit Insurance Corporation
(FDIC)).
22
pool size limit, many jurisdictions impose an OC minimal ratio, which serves as a tool both to increase
MCB quality and to prevent the issuers from increasing MCB issuance.
44. To address the potential increase of fiscal liabilities in the context of the deposit guarantee
scheme, increasing the MCB issuer’s contributions to the BFG may be considered. In the situation where
every issuer (and its MCB program) is well known and licensed by the regulators, the increased payments
can be calculated on a straightforward basis per issuer and be adjusted periodically according to the
bank’s capital and asset and liability composition. Of course, the bank insolvency law and related
practices need to take into account the potential presence of the MCB cover pool, both from the
segregation perspective and to ensure that MCBs do not accelerate. Also, such differentiation of BFG
contributions should not create market distortions by skewing the funding mechanism choices, and they
should not unduly increase the MCB funding costs as well.
45. Finally, the Polish authorities are encouraged to consider that the challenges of asset
encumbrance-related policy, supervision, and practice are subject to an active global policy and
regulatory dialogue. For the market stakeholders in Poland, the process of modernizing their MCB
framework presents an opportunity to utilize the latest global experience on an important issue.
Covered Bonds Issuance Models
46. In Europe, the different Covered Bonds systems can be grouped in 4 issuance models:
MODEL 1: DIRECT ON-BALANCE ISSUANCE RING-FENCED POOL
The MCB issuer is a universal credit institution, either with a qualified covered bond license (e.g., Austria,
Denmark, Finland, Germany since mid-2005, Iceland, Latvia, Slovenia, Sweden) or without such license (e.g.,
Bulgaria, Czech Republic, Greece, Lithuania, Portugal, Spain, Slovakia).
The issuer originates and services both MCB-eligible and non-eligible mortgages. Pool cover loans are MCB-eligi-
ble. MCB issuance is governed by a special legal framework. The cover pool assets are kept on the balance of the
issuer and are segregated (ring-fenced) from the insolvency estate of the credit institution via the creation of a
register that contains a list of all the assets that are part of the cover pool on which MCB holders have a
preferential ranking claim in case of the insolvency of the issuer. The advantages of this model are that it is a
simple structure and assets do not need to be transferred or sold to a different entity. It allows for direct double
recourse but it may require a change in insolvency law in order to avoid issues with other creditors.
MODEL 2: SPECIALIST ISSUER
The covered bond issuer is a separate licensed specialized credit institution. In general, the activities of these
institutions are limited to MCB issuance to fund the cover pool assets. This model exists in France, Ireland,
Norway, Finland and Sweden. MCB issuance is governed by a special legal framework.
The origination and the servicing of the eligible assets and the management of the MCB issuing institution are
done by the parent bank which in general fully owns and supports the issuing entity. Most of the activities of the
covered bond issuer are outsourced to its parent bank. The insolvency segregation of the covered bond issuer from
the parent bank is fundamental. This means that no ring fencing of the cover assets is needed as all the assets of the
issuer are part of the cover pool. The advantage of this structure is that assets are separated in a different legal
entity. The disadvantages are that transfer of the assets is required and there is no automatic direct recourse to the
parent of the issuer which is the originator of the assets (only via consolidation).
MODEL 3: DIRECT ISSUANCE WITH SEGREGATED POOL
The MCB issuer is using a Special Purpose Vehicle (SPV) to achieve insolvency segregation of the cover assets.
23
This structure is used for example in Italy, Netherlands and UK.
The MCBs are issued by the originating bank itself. The cover assets are transferred (or sometimes pledged) to a
legally separated entity, mostly an SPV. This SPV guarantees the payment of the principal and interest of the
Covered Bonds issued by the bank. This model uses securitization techniques to transfer the legal title on the
assets. This model was used in countries where MCBs were initially set up using the general law as no covered
bond law existed. The laws in these countries have basically endorsed those structures by making them law-based
without changing the set-up.
MODEL 4: POOLING
The originator and covered bond issuers are different legal entities, with the issuer aggregating eligible mortgages
from a number of originators. Pooling MCB models exist in legislation and/or in practical use in Austria, Denmark,
Finland, France, Germany, Hungary, Norway, Spain and Switzerland. This option has been used by smaller
originators who combined pools of mortgages with other issuers (typically from other regions) for economies of
scale.
47. In this regard, the Polish authorities are advised to consider that the choice of a particular model
is both a dynamic process (e.g., Germany and Denmark have recently modernized their systems) and in
large part predicated on historical circumstances. Often an emulation of a particular country experience is
not possible due to the fact that present systems are a result of a long and challenging experience. At the
same time, each country may do well by adapting the principles and elements of a given MCB model to
one’s specific circumstances. In this regard, Polish experience with establishing specialized mortgage
banks is telling, as that framework may have put these banks at a disadvantage vis-à-vis large universal
mortgage lenders. This may have led to the current situation where “the portfolio holders cannot issue”
and the “issuers do not have the portfolio”. Unfortunately, making substantial changes to the current
model has become difficult as time has passed – both politically and operationally.
48. The Polish authorities are recommended to adopt Model 1 – and transition from the current
Model #2. Additionally, as the FSAP Mission does not suggest elimination of the specialist mortgage
banks in the near term, elements of the pooling Model #4 can also exist, as the smaller originators may
find it financially and economically efficient to utilize the mortgage banks as the issuing agents. Finally,
the Polish authorities are reminded that regardless of a particular model the high quality of the MCB is
treated by the investor and regulator community on the same level. In other words, since MCBs will be
defined in a single law (modified current 1997 Law), regardless of the issuing entity, MCB legal and
regulatory treatment, as well as broad risk profile and asset characteristics, should be the same.24
49. The Mission recognizes the important and active current industry efforts to optimize the
operations of the mortgage banks vis-à-vis the lenders and appreciates the political and other constraints
which may reduce the regulator flexibility in modifying the existing mortgage covered bonds system. At
the same time, it should be noted that the currently existing Model #2 may not be financially efficient
even if loan transfer obstacles are overcome. MCBs have a dynamic asset pool, i.e., the mortgages in the
cover pool are routinely replaced by the MCB issuer; there would be a need for ongoing whole loan sales
from the mortgage originators to the mortgage banks which are MCB issuers. In addition to the costs, the
risk transfer aspect of such sale transactions would need to be ascertained in the context of their impact on
the MCB investor perception, funding costs, lender incentives, and true dual MCB coverage.
24 Note that of course the terms and conditions of each issuance may differ in tenor, duration, currency, size, etc. Furthermore, the
credit rating of the MCB will be different as they are dependent on the rating of the issuer.
24
Elements of the Modern Securitization Framework
50. In addition to modernizing the MCB environment, the Polish authorities are advised to consider
implementing a legal and regulatory framework for mortgage securitization. This initiative may be a
medium-term undertaking, although certain elements of the system are already in place, e.g., existing
securitization structures for delinquent mortgages and a draft of the securitization law. Furthermore, the
current global regulatory and policy dialogue on securitization is very active. Since Poland does not
currently have a securitization framework in place, the Polish authorities may implement global best
practices without the need to alter an existing system – a potentially easier undertaking compared to the
countries with large securitization markets, e.g., UK and US.
51. RMBSs are structured debt instruments that transfer cash flows from a pool of mortgage loans to
capital market investors who purchase tranches of such securities. RMBSs are created via a series of
transactions which move the mortgages from the balance sheet of the originator to a balance sheet of
another company (Special Purpose Vehicle – a trust or corporation with a narrowly and specifically
defined corporate charter), which does the primary placement of RMBSs.
52. RMBSs have been widely used in the US and other countries, such as Spain, the UK, The
Netherlands, Belgium, Italy, Australia, France, and Japan. Relative to mortgage lending volumes, the
share of funding provided by RMBSs varies by country, but commonly is between 10% and 60% in larger
markets. In 2009 nearly 19% of the outstanding US book of real estate and consumer credit loans worth
USD 18 trillion was funded by private label securitization. A number of countries have also enacted
specific legal and regulatory RMBS frameworks with a view to establish a high-quality instrument
suitable for institutional long-term investors, such as pension funds and insurers. In Eastern Europe,
RMBSs have been used in Russia (ongoing), Kazakhstan and Ukraine (pre-2008).
53. In the context of emerging economies with young mortgage markets, the benefits for lenders and
investors were particularly pronounced. Mortgage lenders typically lack balance sheet or capital strength
RMBS Benefits for Originators and Investors
Originators Investors
Funding diversification. RMBSs provide a stable and low cost source of financing
and allow greater access to credit markets. They reduce lenders’ reliance on retail
deposits and issuance of unsecured commercial debt. They allow smaller, un- or low
rated institutions to access capital markets based on the credit quality of the mortgages
they originate - to get financing at rates appropriate for ‘AAA’ rated firms.
High credit quality instrument (for
senior tranches), portfolio
diversification, and attractive yields
relative to instruments of comparable
credit quality.
Risk transfer. RMBSs transform illiquid mortgages that otherwise would be held in a
bank’s portfolio, into marketable securities. Issuance of RMBSs is one means of
transferring credit, liquidity, interest rate, prepayment and market risk associated with
that collateral to investors. Ability to achieve balance sheet asset derecognition varies
by jurisdiction.
RMBS investors could avoid
exceeding concentration limits (both
regulatory and internal limits) on
exposures to a single name.
Revenue generation. RMBSs have been a means for generating revenues, e.g.,
origination fees, underwriting and structuring fees, selling RMBS credit and liquidity
enhancements. Issuers also created revenue streams through credit arbitrage – the
positive spread differential between longer-term assets and shorter RMBS bonds.
RMBSs facilitate portfolio risk
management as holding securitized
assets may have a low correlation
with other portfolio components.
Regulatory capital and financial reporting. Removal of long-term mortgages from
the balance sheet improves financial ratios, such as the LCR or ROA, and reduces the
balance sheet duration gap, exposure to capital provisioning, and reserves in case of
mortgage portfolio performance deterioration.
RMBS risk-adjusted returns are
typically higher relative to a similarly
rated sovereign debt, which leads to
higher returns per rating.
25
to carry significant duration gaps created by increased mortgage loan portfolios. Plain vanilla RMBSs
were viewed as a relatively simple and safe mechanism to introduce a low risk asset class to local
institutional investors which lacked diversification in private fixed income instruments. International
investors were also interested in the arbitrage between performance of “cherry picked” mortgage loans in
the collateral pools and perceived high legal, country or transactional risks which skewed RMBS
yield/risk performance vis-à-vis other debt instruments.
54. Since 2007 the global private mortgage securitization markets have been largely closed due to
lack of investor interest in the asset class. Although origination volumes have dropped in light of the
global financial crisis, high unemployment, and economic slowdown, without RMBSs lenders have fewer
options in obtaining long-term capital market funding. In ECA this is relevant for countries where RMBS
issuance was taking place before the crisis – Russia, Kazakhstan and Ukraine. Although in Russia
securitization has been continuing, in other countries it has ceased. As a consequence, borrowers are faced
with inefficient mortgage product features and high pricing.
55. RMBSs, particularly in developed markets with a large volume of complex products, were
affected by misaligned incentives or conflicts of interest. Certain market idiosyncrasies may have
facilitated such misalignment, e.g., the evolution of the originate-to-distribute model, the involvement of a
large number of parties in transactions, complex yet opaque investor disclosure and transaction
documentation, and not easily deducible risk path between loan originators and investors. Issuer and
lender compensation programs, which emphasized volume and growth, overshadowed concerns about the
quality of the mortgages. Investors, regulators, and rating agencies came to rely heavily on the
representations and warranties made by originators. Investors chose to respond to growing product
complexity by relying on credit ratings rather than conducting appropriate due diligence. Lenders had
incentives to choose riskier assets in constructing asset pools. On the investor side, portfolio managers
and hedge funds were incentivized to maximize short-term gains and yields without considering long-
term risk. Investors failed to assess the RMBS risks adequately in part due to the information asymmetry
which tended to favor the supply side and in part due to own institutional capacity constraints
56. However, RMBS and MCB frameworks are complementary, as the products have different
characteristics and appeal to different investor needs. From the mortgage lender perspective, loans
compliant with RMBS and MCB may also have different terms and conditions. As RMBSs’ perceived
quality is more closely linked with the performance of the mortgage loan pool, smaller and lower rated
originators have a natural incentive to utilize RMBSs instead of MCBs.
57. The Polish authorities are encouraged to consider that RMBSs, in addition to MCBs, will likely
improve the overall risk profile of the national mortgage portfolio by imposing stringent loan eligibility
criteria, in addition to requiring improved practices from the market participants. Specific RMBS benefits
in this regard compared to MCBs include:
a. More advanced, detailed and transparent loan level investor and public disclosure – both
at the time of securitization and ongoing;
b. Strong links between performance and pricing of the mortgage pools and the RMBSs
which incentivize originators and servicers to follow prudent practices;
c. Ability, due to the structured nature of RMBSs, to reach different investors with one
issuance, as different tranches are purchased by different investors;
d. A viable capital market funding channel for smaller, lower rated originators,
incentivizing them to follow high quality lending and servicing practices;
26
e. The RMBS requirement for significant standardization of loans and market practices may
serve as a driver for the whole market to improve;
58. Additionally, the Polish authorities are encouraged to consider that RMBSs allow lenders to
deleverage their balance sheets, which may be desirable in the context of a large share of mortgages in the
banking loan book as well as a potentially highly risky profile of the aggregate portfolio. Admittedly,
RMBS transactions with the most problematic mortgages, e.g., LTV > 100%, may be challenging for the
issuers, although their creativity should not be underestimated. KNF is advised to consider establishing
minimum loan eligibility criteria for the domestic RMBS, e.g. limiting LTV to [80% without and 90%
with appropriate MI] on the pool and individual loan level, requiring performing PLN loans, etc.
59. As the other countries with previously large securitization markets, including the US and EU are
actively searching for the legal and regulatory frameworks that would instill investor confidence in
RMBS, the Polish authorities are advised to implement such best practices in setting up the national
mortgage securitization environment.
The main themes of post 2008 global initiatives to re-start the RMBS markets
Re-aligning incentives of transaction parties, e.g., originator risk retention, rating agency governance improvements, and
reduced reliance on credit ratings.
Improving quality, knowledge of, and monitoring of the mortgage assets, e.g., loan level disclosure, “qualified mortgage loans”,
strengthened auditing and due diligence.
Simplifying and standardizing RMBS structures, e.g., improved transaction documentation, uniform definitions
Risk retention
and alignment
of incentives
In the EU, Article 122(a) of the Capital Requirements Directive (CRD II) includes a minimum risk retention
rate … which shall not be less than 5% of the total issuance. Similar risk retention requirements will be
included in forthcoming Directive 2009/138/EC known as Solvency II.
In the US, Section 941 of the Dodd-Frank Wall Street Reform and Consumer Protection Act (the Dodd-
Frank Act) requires a securitizer to retain at least 5 % of the credit risk…. The “safe harbor” provisions of
the FDIC Securitization Rule impose a 5 % credit risk retention requirement for bank-sponsored RMBSs.
Pools consisting of high quality Qualifying Residential Mortgages may be excluded from such requirements.
Transparency
and disclosure
In 2009 IOSCO issued guidance on increasing transparency of risk verification and assurance practices and
improving asset pool performance information available to investors on an initial and ongoing basis. In the
EU, new disclosure requirements (2010 CRD amendments) require that prospective investors have readily
available access to all materially relevant data on securitization structures. The ECB and the Bank of
England have launched initiatives to implement new disclosure requirements in the context of collateral
eligibility frameworks. The first EU loan-by-loan RMBS template was published by the ECB in 2010.
In the US, the Dodd-Frank Act has provisions relating to disclosure for ABS issuers. In 2010, the SEC
proposed revisions to “Regulation AB”, which included new requirements to increase the transparency and
standardization in the private ABS market. In Japan, the supervisory guidelines for securities companies
were revised in order to ensure the traceability of underlying assets of securitized products in April 2008.
Credit agency
governance
and regulation
In Japan, in line with IOSCO’s revised code of conduct (2008), rating agencies are required to publish
information that may be deemed valuable in an assessment by a third party of the appropriateness of the
credit rating. In the US and Europe, rating agencies will be subject to increased disclosure requirements to
increase transparency in connection with structured finance ratings.
Banking
regulation
Basel III includes elements that will significantly affect the incentives for banks to securitize loans and invest
in RMBS, in part via significantly increased RMBS risk weights. To address the lack of investor due
diligence and to deter them from relying solely on external credit ratings, the Basel framework requires
banks to meet specific operational criteria in order to use the risk weights specified in the Basel II
securitization framework. Solvency II will also establish new capital requirements for the insurance sector
with increased risk sensitivity and make investment in RMBSs potentially less attractive to insurers.
27
60. An important aspect of restarting private RMBS markets in any country (or initiating mortgage
securitization in countries like Poland, where it does not exist currently) is creation of the “Label”, i.e., a
high quality framework of securitization. Establishing a RMBS Label is a complex initiative with
elements in virtually all aspects of mortgage securitization. The Label brand name recognition and value
should be such that its potential loss would be a deterrent to any issuers from being lax on origination,
servicing and provision of transparency. It is also important that introducing a Label allows for
simultaneous (i) strengthening of existing statute or law, e.g., timely deliverance of information, as well
as (ii) promotion of improvements before they are codified, e.g., trade transparency, loan servicing
pooling and quality of service agreements.
61. RMBSs certified to carry the Label can be considered for certain regulatory preferences,
particularly in terms of investor portfolio management and capital allocation. Investor confidence in
RMBS has severely diminished and a re-start would require increasing actual and perceived quality of the
instrument. Given that RMBS instrument performance relies on the performance of the underlying assets,
the quality of the mortgage loans and availability of information about such quality are paramount.
62. Specific Label related practices include:
a. Eligible mortgage loans should be of high, consistent and verifiable quality. Collateral
quality should be based on auditable information and processes; it should be perceived as
high by investors and rating agencies. The quality of the mortgage pool should be
credibly and transparently assessed and certified before issuance as part of enhancing
quality of the pre-issuance process. Investors should be able to perform, should they so
wish, certain verifications. Pre-issuance audits should pertain to the conformity of loan
files – in terms of data and loan quality – to the representations made by transaction
sponsor. Current, complete and accurate data on mortgage loans is vital and should be on
loan level and updated at least monthly. Loan-level trustee reports during the life of the
Label RMBS instrument and data on the pool performance should be available from
origination of individual mortgage loans.
b. Legally established limits on LTV and DTI. Additionally to actual numeric limits, the
quality of assessing the ratios is important and raises the issues of real estate appraisal
industry, income verification and overall underwriting policies and practices.
Standardized terminology is critical, as RMBS transactions may involve multiple loan
Private RMBS Restart Agenda
1. Assess the market thoroughly, as not all global practices and issues may apply.
2. Achieve high quality in the Asset, Instrument and Market.
3. Convey quality with transparency, disclosure and standardization.
4. Introduce a Label to reflect the holistic nature of the changes and to further elevate the RMBS quality.
What is RMBS Label Quality?
Mortgage Asset Quality, i.e., the degree of standardization of transaction documentation, terminology, underwriting and
servicing practices, foreclosure regime and, availability of loan-level analytical data.
RMBS Instrument Quality, i.e., trustee quality and functionality, predictability and transparency of transaction documents,
initial and ongoing investor disclosure, servicer regime, including the framework and rules for servicer substitution.
Market Operations Quality, i.e., transparency of primary and secondary trading in price formation, transparency of movements,
incentives for issuance participants, including rating agencies, RMBS investment regime vis-à-vis other instruments.
28
originators, i.e., all market stakeholders should have a common understanding of key
performance and analytical notation, such as delinquency, LTV, and servicer.
c. Label RMBS structures can be standard and plain vanilla, possibly 3-tier – senior,
mezzanine and equity. This facilitates external credit enhancement mechanisms and
allows for straightforward risk retention by the originators, caters to different types of
investors, and eases analytical modeling and pricing. External Label RMBS credit
enhancement features, e.g., guarantees, liquidity registers, etc., should be transparent and
standardized so that investors have the ability to evaluate their impact on the credit
quality of the transaction. Legal agreements used in Label RMBS transactions should be
standardized and created using “a by reference” model, i.e., pooling and servicing
agreements, whole loan sales agreements, servicing and special servicing documentation,
etc.
d. Investors and market participants should have unhindered and free access to accurate,
timely and complete RMBS performance information. This includes loan level at
securitization static pool data, periodic loan level disclosure, and investor reports with
relevant information.
63. RMBS transactions have an impact on financial markets and mortgage funding only in case there
is a sufficient volume of primary and secondary trading. Market infrastructure should support transparent
price formation and absence of collusion at issuance. Additional market operational quality features worth
considering are:
a. A backup servicer should be provided for in the transaction documentation including
scope of its services and remuneration.
b. While market making in relation to Label RMBS may be impractical for all tranches,
“benchmark/reference tranches” of particularly high quality would benefit from it.
c. Statutory portfolio allocations of key institutional investors should be de-linked from
ratings per se, and instead should include quality and instrument type guidance.
64. Particularly relevant to mortgages, whole loan sale and purchase transactions should be subject to
a neutral legal and regulatory framework, particularly in such aspects as the taxation regime, rights re-
registration process, servicing transfer, disclosure, and transparency. In this regard, SPV establishment
and operations play an important role, as economic and financial efficiency of whole loan transfers to
such companies as part of a securitization transaction critically affect the Label securities.
65. Notwithstanding limited presence of this instrument in ECA, RMBSs can play an important role
in the mortgage industry’s spectrum of long term wholesale funding. While all of the above discussion is
relevant for emerging markets, the initial step that market stakeholders could undertake is to thoroughly
assess the deficiencies in the existing RMBS framework. The overall goal of such an assessment would
be to identify areas of securitization policies and practices that may require changes along the lines
discussed above.
66. From the implementation perspective, the development of the MCB legal and regulatory
framework, as suggested in the previous section, does not need to coincide chronologically with the
development of the mortgage securitization framework. Passage of legal and regulatory amendments,
especially numerous ones under a single comprehensive agenda, may require significant political
discussions, which may not be relevant to the substance of such changes. The Polish authorities are
29
encouraged to ascertain and devise the most expedient mechanism of supporting the adoption of both
components of the capital market funding mechanisms for the mortgage industry.
Conclusions
67. The Polish Mortgage market, while well developed and large, presents several opportunities to
the stakeholders for significant strengthening. The existing portfolio contains a large portion of highly
risky assets which may present a micro and macro stability challenge, should its performance deteriorate.
In this regard the authorities are encouraged to actively cooperate with the industry in performing a
comprehensive and thorough data collection and analysis exercise with the goal of identifying portfolio
segments of the potentially risky mortgages on the aggregate and institutional level. Furthermore, should
the analysis so warrant, a loan modification program may be designed to proactively reduce the credit
risks for the borrowers and lenders.
68. In addition to the activities aimed at the existing portfolio, the Polish authorities are encouraged
to strengthen the existing regulatory environment by revising Recommendation S is a few areas, such as
designing modern guidance on the underwriting practices, usage of mortgage insurance and establishing a
matrix of key loan parameters which can be used to compensate for the risk factors, and not to allow them
to pile one on top of the other. Such guidance should be reflected in lender capital allocation policies.
69. Furthermore, the Polish authorities are encouraged to utilize a modern well-developed capital
market access channels to both strengthen the banking sector in terms of funding diversification and
reduction of liquidity gaps and create additional mechanism to ensure prudent origination and servicing
practices. In this regard modernization of the MCB framework by allowing universal banks to issue
Mortgage Covered Bonds and establishing a mortgage securitization framework are advised.
30
Annex 1. MCB in the Legal, Regulatory and Marketplace context
Covered Bonds are debt securities issued by a credit institution that are backed by a dynamic cover pool
of high quality assets. Investors have double recourse to the issuer and to the cover pool. MCBs are
mostly issued under a dedicated legal framework and the issuers are publicly licensed and supervised by
the regulator. The covered bond is one of the key components of European capital markets. The asset
class plays an important role in the financial system by contributing to the efficient allocation of capital
and ultimately economic development and prosperity. With over EUR 2.7 Trillion outstanding in the EU
at the end of 2012, Covered Bonds continue to play an essential role in bank funding strategies for the
European banks – the amount of outstanding mortgage Covered Bonds is equivalent to around 20% of
outstanding residential mortgage loans in the EU.
The introduction of a MCB system represents a strategically important addition in the funding options
available to mortgage lenders. This on-balance-sheet instrument, based on specific legislation and
supervision, provides the market with a long-term funding tool with cost-efficient performance on the
issuer’s side and a stable and safe long-term, liquid investment on the investor’ side, contributing
significantly to the creation of an efficient housing market, capital market development, and financial
stability. In view of the main features of MCB, the smooth functioning of this market is perceived by
regulators as an important contribution from a financial stability perspective and as a priority in current
debate on Basel III and revision of the European regulatory framework, resolution regimes and the REPO
policies.
Under EU law there are two working definitions of Covered Bonds. Bonds which meet these criteria are
eligible for preferential treatment in various ways (including a lower risk weighting, exemption from
certain concentration rules and eligibility for bank liquidity ratios). Furthermore, the European Central
Bank’s (ECB) rules defining eligible assets for repo operations refer to definitions under EU law.
The two definitions of Covered Bonds are those contained in the Directive on Undertakings of Collective
Investment in Transferable Securities (UCITS Directive or EU directive 2009/65) and in the Capital
Requirements Directives (CRD or EU Directives 2006/48 and 2006/49):
Article 52(4) of the UCITS Directive requires a special regulatory regime designed to protect
the interests of covered bondholders to be in place:
CBs must be issued by an EU credit institution which is subject to special public supervision
by virtue of legal provisions protecting bondholders;
Bondholders’ claims on the issuer must be fully secured by eligible assets until maturity;
Bondholders must have a preferential claim on the cover assets in case of the issuer’s default;
The cover assets must be constituted only of assets of specially defined types and credit
quality. The CRD (Annex VI, Part 1, Paragraph 68) defines a list of assets eligible to back the
bonds;
New quantitative restrictions must be imposed on cover assets (e.g. 15% exposure on credit
institutions);
The issuers of MCBs must meet certain minimum requirements regarding mortgage property
valuation and monitoring.
Key elements to be considered in the MCB legislation are:
31
Segregation and ring-fencing of the cover pool in case of issuer insolvency: procedures
should clearly state that in such event MCB holders have preferential claim on the assets in
the cover pool which should survive and be fully segregated from the bankruptcy estate;
Licensing requirements: Each issuer needs to apply for a specific license by the regulator.
The license should be public and monitored on an ongoing basis. Conditions for licensing
should include regulatory capital levels and other limitations, e.g., a specified minimum level
of core capital, adequate technical resources, and a sufficient organizational structure;
Strict public supervision on issuer and cover pool: The regulator can outsource reporting to a
Cover Asset Monitor which represents the interests of the bondholders and if required acts to
ensure timely payment by procuring liquidity, acting as a servicer, sell or refinance assets, or
issue new bonds.
Strict eligibility criteria for cover assets: Residential or Commercial mortgages or public
sector exposures are in general the only types of eligible assets. Mortgages are subject to
strict LTVs [up to 80%] and valuation conditions. Substitution assets are limited to 15% and
subject to strict eligibility conditions. Derivatives can be part of the pool but should be clearly
defined. Each class of assets should be in a separate cover pool.
The majority of the Covered Bonds issued in Europe are UCITS-compliant and in most cases also CRD-
compliant which means that the capital weights for bank investors are much lower. Under the Basel I and
Basel II standardized approach, the capital weight for UCITS-compliant MCBs is 20% and for CRD-
compliant MCBs 10%. The capital weight under Basel II advanced approach can be lower but depends on
the issuer rating. MCBs continue to be attractive for bank investors as they are eligible as level-2 high-
quality liquid assets for the Liquidity Coverage Ratio (LCR) under Basel III. This favorable regulatory
treatment has resulted in some Asset Backed Securities (ABS) investors (especially Asset Liability
management (ALM)) desks of banks for their liquidity portfolios) turning to the MCB-market in Europe.
MCBs in 5-year maturities are dominated by bank investors while the largest portion of MCBs issued in
longer maturities are taken up by insurers and pension funds. Central banks account for 15-20% of the
issuance. Covered Bonds are also treated favorably as repo-eligible collateral by the ECB and other
central banks:
a. Covered Bonds only need a BBB- rating while initially only AAA rated-ABS were
eligible; ABS needed two ratings while MCBs with one rating are eligible;
b. The haircut applied to CBs (0.5-7.5%) is much lower than for ABS (minimum 16.4% on
market value).
32
Annex 2. Comparison between MCBs and RMBSs
Feature Comparison between MCBs and RMBS
Covered Bonds RMBS
Issuer Licensed credit institution, in most cases a bank or
dedicated specialist issuer SPV
Legal Framework Generally specific law and regulation
Supervision Bank supervision, trustee, capital market authority Bond trustee, capital market authority
Payments from Lender - operating cash flows until insolvency, then
priority claim on cover pool
SPV – from collateral pool, tranche waterfall
defines priority between bondholders
Payment
acceleration Upon issuer and pool default On specific events by triggers
Claim
Investors have dual claim: recourse on issuer and
priority claim on the cover pool assets in case of issuer
insolvency; claims of bond holders rank pari passu and
at same level with swap counterparties prior to all
other creditors
Investors have claim on collateral and its cash
flows; priority to other bond holders depends
on tranche
Balance sheet
treatment Assets mostly remain on issuer’s balance sheet
Assets (or associated risks) are transferred to
an SPV
Pool Homogeneity High degree of standardization, asset eligibility defined in law
Cover pool Dynamic & managed Mostly static, asset substitution may be
possible in certain cases
Credit
enhancement
Overcollateralization provided by issuer. Min OC
required by law
Various, including OC, reserves, guarantees,
structuring, etc.
Issue structure Mostly bullet (or soft bullet), independent of asset
amortization
Pass-through, according to asset amortization
and issue structure
Interest rate Mostly fixed Typically asset dependent
Documentation Program documentation Documentation per transaction
Security All series of bonds have recourse to one single cover
pool Each issue is secured by distinct asset pool
Capital weighting
Typically 10% in standardized approach for CRD-
compliant CB (20% for UCITS-compliant bonds),
(new jurisdictions may start at 20% for CRD purposes)
Depends on rating and jurisdiction
Credit risk Retained by issuer Passed to investors, may flow to lower
tranches retained by transaction sponsor
Servicing By issuers Typically by specialized servicers
Valuation and LTV Eligibility criteria vary, often defined by law
Disclosure Minimum level defined by law; extensive defined by regulation or marketplace
ALM Defined by law or program Typically pass-through
Rating
methodologies
Separate methodologies combining fundamental and
structured finance elements. Issuer rating forms
minimum rating and basis for analysis. Maximum
rating dependent on country, legal framework, pool
quality, ALM and liquidity risk.
Multifaceted: structured finance approach +
servicer ratings. Maximum rating dependent
on country, legal framework, pool quality,
ALM and liquidity risk.
Liquidation Insolvency remoteness of cover pools Insolvency remoteness of SPV
While public issuance of RMBSs recovered in the past two years, issuance levels are still far from
historical highs. In the case of MCBs on the other hand, issuance levels were at historically high levels in
H2 2010 and Q1 2011, although issuer retention (i.e., bonds not placed to market) remained high at over
75% in the EU. There are key differences between the two products:
33
a. RMBS investors are exposed to the risk of underperformance of the cover pool. During
the financial crisis, high non-performing loans and lower pre-payments were drivers of
cover pool underperformance and maturity extensions of RMBS;
b. MCBs are bank bonds, and holders of MCBs benefit from a preferential claim on a cover
pool, the support of the issuing bank and every kind of external support provided to the
issuing bank;
c. MCB pools are dynamic, and due to typically high asset liability mismatches between
cover assets and outstanding MCBs, OC requirements by rating agencies for MCBs are
much higher than credit enhancements for senior tranches of RMBSs, in turn increasing
investor protection.
MCBs are an on-balance sheet funding tool; RMBSs free up capital. In contrast to securitization, in the
MCB context mortgage assets remain on the balance sheet of the issuers and all credit risk is retained by
the issuers. Some structures could be seen as utilizing a quasi-SPV specifically dedicated to the issuance
of MCBs because although the issuer is a licensed credit institution, it is in fact a specialized Covered
Bond bank. The specialized issuer uses the bond placement proceeds to buy mortgage loans at the lender
bank or to grant loans to the operating bank, the originator of the mortgage loans. In case of the latter, the
operating bank keeps the mortgage loans on its balance sheet and pledges them as security for the loan
received from the Covered Bond issuing bank. However, in both cases, Covered Bonds are an on-bank-
balance sheet funding tool and as such have no asset derecognition or capital relief consequences for the
issuer.
MCB are backed by a dynamic cover pool; RMBSs are backed by a static pool. All outstanding bonds
issued under one MCB program by an issuer are typically backed by all loans in the cover pool,
independent on the timing of the issuance. The pool is dynamic and managed which means that issuers
can add and delete assets from the cover pool as long as all legal and rating agency criteria are met. The
issuer typically takes out non-performing loans (i.e., keeps the pool clean). In most countries, issuers are
obliged to do so by law. Increasing levels of non-performing loans in the cover pool could indicate that
the issuer is no longer able to support the cover pool. In case of issuer insolvency, the pool in general
becomes static and no further assets will typically be added to the cover pool and no further Covered
Bonds can be issued. The level of support to CB investors can be increased over time by higher OC
levels. RMBS investors are generally more exposed to the performance of the pool, depending on bond
structure and credit enhancement mechanisms, as well as on which tranche was purchased.
The reason for the higher credit enhancement of MCBs comes from the maturity mismatch between the
bonds (mostly bullet structures) and cover pool assets (amortizing). The resultant MCB liquidity risk can
be mitigated by including ALM tests and/or a higher OC required by the rating agencies. RMBS notes are
in general amortizing on a pass-through basis with the amortization of the portfolio and generally have no
liquidity risk. Additionally to the credit risk of the pool assets, risks are
a. the potential lower yield of newly added assets (negative carry risk of differing amortization
profiles of bonds and pool assets);
b. interest rates and currency mismatches between fixed rate MCBs and (often) variable rate
mortgage loans, and
c. the need to sell cover assets in a short period of time in case of issuer insolvency to pay
MCBs with bullet maturities.
34
MCB recourse is to the issuer and cover pool; RMBS recourse is to the asset pool. An important
difference between MBC and RMBS is that MBC holders have double recourse, firstly to the issuing
bank and secondly to the cover pool of assets while securitization investors only have recourse to the pool
of assets in the SPV. Some view that the benefit from the additional recourse to the bank is limited for
specialist MCB issuers who fund most of their balance sheet via MCBs due to potential correlation
between the quality of the cover pool and the credit quality of the issuer.
MCB issuers are licensed and supervised by regulators so benefit from intrinsic support in most cases.
Regulators often monitor the impact of the issuance of MCBs on the issuer and its unsecured creditors
and depositors. In some cases, regulators have set maximum limits to the amounts of MCBs that issuers
can issue without specific permission of the regulators. Moreover, regulations relating to liquidity buffers,
leverage limits, reserve requirements and valuations are likely to make banks fundamentally stronger
which in turn would support MCBs. The regulatory support and oversight is seen as one of the reasons for
MCBs outperforming RMBSs during the financial crisis.
One of the main risks of highly rated RMBSs is maturity extension as the repayment profile depends on
the (p)repayment rate of the underlying assets which can vary over time. Falling prepayment rates along
with the lack of fully functioning debt capital markets has meant extension risk has become a core
consideration in European RMBSs. MCB OC is typically higher than subordination in RMBS. Typically,
OC requirements of rating agencies to achieve AAA ratings are much higher for MCBs than for senior
RMBS tranches. This is mainly due to MCBs facing not only credit risk but also market risks, especially
liquidity risk stemming from maturity mismatches between cover pool assets and outstanding fixed bullet
MCBs. MCB legislation in many countries stipulates a minimum level required by law (2% in France and
Germany, 20% for Spanish Cédulas, and 5% in other countries).
On an even playing field, MCBs and securitization should be viewed as complementary rather than
competing funding vehicles. During normal times, they both increase the range of available financial
products, benefiting borrowers, financial intermediaries, and savers. During periods of market stress,
MCBs provide a time-tested funding backstop, albeit mainly for investment-grade banks (rated “BBB-
/Baa3” and better). Moreover, the legally defined quality principles for selection of underlying assets and
strict segregation of assets of the cover pool suggest little valuation uncertainty even in distress situations.
In regard to securitization, the multitude of complex relationships between issuers, arrangers and liquidity
providers in securitization transactions could introduce an element of uncertainty in times of stress, such
as the adverse impact of originator insolvency (despite the insulation of the reference portfolio of
securitized assets).
Generally, particularly post 2007, RMBSs face increasing legal and regulatory restrictions. On the other
hand, legislators and regulators increasingly support MCBs. For instance,
a. CRD II – 5% retention and greater disclosure requirements for RMBSs,
b. CRD III – more onerous capital requirements for securitizations held in trading
books,
c. CRD IV – more onerous liquidity requirements,
d. Solvency II – capital requirements for insurance companies and credit rating agency
legislation,
e. ECB collateral requirements – two AAA ratings in case of RMBS compared to BBB-
in case of MCBs.